Smart Sensor-Assisted ANN-Based Distributed MPPT Framework for Enhanced Energy Harvesting in Partially Shaded Photovoltaic Systems

Partial shading conditions (PSCs) significantly degrade photovoltaic (PV) system performance by creating multiple local maxima and reducing extractable energy. This paper presents a distributed Maximum Power Point Tracking (MPPT) approach that uses a TotalCross-Tied (TCT) configuration of String Integrated Converters (SICs) to improve power harvesting under varying shading conditions. Three MPPT algorithms, namely Perturb and Observe (P&O), Incremental Conductance (I&C), and Artificial Neural Network (ANN), were evaluated through MATLAB/Simulink simulations under several irradiation patterns: uniform irradiation, uneven row shading, uneven column shading, and elongated narrow shading. The framework also incorporates sensor-assisted monitoring of voltage, current, irradiance, and temperature to supply real-time operating information to the MPPT controller. This distributed sensing supports the ANN’s adaptation to changing environmental conditions. Simulation results show that the ANN-based MPPT achieved the highest output power of 6003 W, with better voltage and current characteristics, under uniform irradiation of 1000 W/m². The ANN also outperformed I&C under uneven column and elongated narrow shading, whereas I&C performed slightly better under uneven row shading owing to its slope-based tracking. These findings indicate that intelligent MPPT methods can improve PV system performance under both uniform and partially shaded conditions, with the best algorithm depending on the shading pattern. The proposed TCT-SIC architecture, combined with intelligent MPPT algorithms and sensor-based monitoring, offers a promising and scalable solution for PV systems operating under partial shading conditions.